[textanalytics] use task name to correlate inputs/outputs for actions

sdk/textanalytics/azure-ai-textanalytics/azure/ai/textanalytics/_models.py

@@ -1409,13 +1409,14 @@ def __repr__(self, **kwargs):
         return "RecognizeEntitiesAction(model_version={}, string_index_type={}, disable_service_logs={})" \
             .format(self.model_version, self.string_index_type, self.disable_service_logs)[:1024]
 
-    def to_generated(self):
+    def to_generated(self, task_id):
         return _latest_preview_models.EntitiesTask(
             parameters=_latest_preview_models.EntitiesTaskParameters(
                 model_version=self.model_version,
                 string_index_type=self.string_index_type,
                 logging_opt_out=self.disable_service_logs,
-            )
+            ),
+            task_name=task_id
         )
 
 
@@ -1480,14 +1481,15 @@ def __repr__(self, **kwargs):
             self.disable_service_logs,
         )[:1024]
 
-    def to_generated(self):
+    def to_generated(self, task_id):
         return _latest_preview_models.SentimentAnalysisTask(
             parameters=_latest_preview_models.SentimentAnalysisTaskParameters(
                 model_version=self.model_version,
                 opinion_mining=self.show_opinion_mining,
                 string_index_type=self.string_index_type,
                 logging_opt_out=self.disable_service_logs,
-            )
+            ),
+            task_name=task_id
         )
 
 
@@ -1546,14 +1548,15 @@ def __repr__(self, **kwargs):
             self.disable_service_logs,
         )[:1024]
 
-    def to_generated(self):
+    def to_generated(self, task_id):
         return _latest_preview_models.PiiTask(
             parameters=_latest_preview_models.PiiTaskParameters(
                 model_version=self.model_version,
                 domain=self.domain_filter,
                 string_index_type=self.string_index_type,
                 logging_opt_out=self.disable_service_logs
-            )
+            ),
+            task_name=task_id
         )
 
 
@@ -1593,12 +1596,13 @@ def __repr__(self, **kwargs):
         return "ExtractKeyPhrasesAction(model_version={}, disable_service_logs={})" \
             .format(self.model_version, self.disable_service_logs)[:1024]
 
-    def to_generated(self):
+    def to_generated(self, task_id):
         return _latest_preview_models.KeyPhrasesTask(
             parameters=_latest_preview_models.KeyPhrasesTaskParameters(
                 model_version=self.model_version,
                 logging_opt_out=self.disable_service_logs,
-            )
+            ),
+            task_name=task_id
         )
 
 
@@ -1649,11 +1653,12 @@ def __repr__(self, **kwargs):
                 self.model_version, self.string_index_type, self.disable_service_logs
             )[:1024]
 
-    def to_generated(self):
+    def to_generated(self, task_id):
         return _latest_preview_models.EntityLinkingTask(
             parameters=_latest_preview_models.EntityLinkingTaskParameters(
                 model_version=self.model_version,
                 string_index_type=self.string_index_type,
                 logging_opt_out=self.disable_service_logs,
-            )
+            ),
+            task_name=task_id
         )

sdk/textanalytics/azure-ai-textanalytics/azure/ai/textanalytics/_request_handlers.py

@@ -6,7 +6,7 @@
 
 
 import six
-
+from ._generated.models import EntitiesTask, PiiTask, EntityLinkingTask, SentimentAnalysisTask
 from ._models import (
     DetectLanguageInput,
     TextDocumentInput,
@@ -80,6 +80,17 @@ def _determine_action_type(action):
         return _AnalyzeActionsType.ANALYZE_SENTIMENT
     return _AnalyzeActionsType.EXTRACT_KEY_PHRASES
 
+def _determine_task_type(action):
+    if isinstance(action, EntitiesTask):
+        return _AnalyzeActionsType.RECOGNIZE_ENTITIES
+    if isinstance(action, PiiTask):
+        return _AnalyzeActionsType.RECOGNIZE_PII_ENTITIES
+    if isinstance(action, EntityLinkingTask):
+        return _AnalyzeActionsType.RECOGNIZE_LINKED_ENTITIES
+    if isinstance(action, SentimentAnalysisTask):
+        return _AnalyzeActionsType.ANALYZE_SENTIMENT
+    return _AnalyzeActionsType.EXTRACT_KEY_PHRASES
+
 def _check_string_index_type_arg(string_index_type_arg, api_version, string_index_type_default="UnicodeCodePoint"):
     string_index_type = None
 

sdk/textanalytics/azure-ai-textanalytics/azure/ai/textanalytics/_response_handlers.py

@@ -200,31 +200,29 @@ def _get_property_name_from_task_type(task_type):
         return "sentiment_analysis_tasks"
     return "key_phrase_extraction_tasks"
 
-def _get_good_result(current_task_type, index_of_task_result, doc_id_order, response_headers, returned_tasks_object):
+def _get_good_result(task, doc_id_order, response_headers, returned_tasks_object):
+    current_task_type, task_name = task
     deserialization_callback = _get_deserialization_callback_from_task_type(current_task_type)
     property_name = _get_property_name_from_task_type(current_task_type)
-    response_task_to_deserialize = getattr(returned_tasks_object, property_name)[index_of_task_result]
+    response_task_to_deserialize = \
+        [task for task in getattr(returned_tasks_object, property_name) if task.task_name == task_name][0]
     return deserialization_callback(
         doc_id_order, response_task_to_deserialize.results, response_headers, lro=True
     )
 
 def get_iter_items(doc_id_order, task_order, response_headers, analyze_job_state):
-    iter_items = defaultdict(list) # map doc id to action results
-    task_type_to_index = defaultdict(int)  # need to keep track of how many of each type of tasks we've seen
+    iter_items = defaultdict(list)  # map doc id to action results
     returned_tasks_object = analyze_job_state.tasks
-    for current_task_type in task_order:
-        index_of_task_result = task_type_to_index[current_task_type]
+
+    for task in task_order:
         results = _get_good_result(
-            current_task_type,
-            index_of_task_result,
+            task,
             doc_id_order,
             response_headers,
             returned_tasks_object,
         )
         for result in results:
             iter_items[result.id].append(result)
-
-        task_type_to_index[current_task_type] += 1
     return [
         iter_items[doc_id]
         for doc_id in doc_id_order

sdk/textanalytics/azure-ai-textanalytics/azure/ai/textanalytics/_text_analytics_client.py

@@ -18,7 +18,7 @@
 from ._base_client import TextAnalyticsClientBase
 from ._request_handlers import (
     _validate_input,
-    _determine_action_type,
+    _determine_task_type,
     _check_string_index_type_arg
 )
 from ._response_handlers import (
@@ -896,32 +896,26 @@ def begin_analyze_actions(  # type: ignore
         continuation_token = kwargs.pop("continuation_token", None)
 
         doc_id_order = [doc.get("id") for doc in docs.documents]
-        task_order = [_determine_action_type(action) for action in actions]
+        generated_tasks = [action.to_generated(str(idx)) for idx, action in enumerate(actions)]
+        task_order = [(_determine_task_type(a), a.task_name) for a in generated_tasks]
 
         try:
             analyze_tasks = self._client.models(api_version='v3.1').JobManifestTasks(
                 entity_recognition_tasks=[
-                    t.to_generated() for t in
-                    [a for a in actions if _determine_action_type(a) == _AnalyzeActionsType.RECOGNIZE_ENTITIES]
+                    a for a in generated_tasks if _determine_task_type(a) == _AnalyzeActionsType.RECOGNIZE_ENTITIES
                 ],
                 entity_recognition_pii_tasks=[
-                    t.to_generated() for t in
-                    [a for a in actions if _determine_action_type(a) == _AnalyzeActionsType.RECOGNIZE_PII_ENTITIES]
+                    a for a in generated_tasks if _determine_task_type(a) == _AnalyzeActionsType.RECOGNIZE_PII_ENTITIES
                 ],
                 key_phrase_extraction_tasks=[
-                    t.to_generated() for t in
-                    [a for a in actions if _determine_action_type(a) == _AnalyzeActionsType.EXTRACT_KEY_PHRASES]
+                    a for a in generated_tasks if _determine_task_type(a) == _AnalyzeActionsType.EXTRACT_KEY_PHRASES
                 ],
                 entity_linking_tasks=[
-                    t.to_generated() for t in
-                    [
-                        a for a in actions
-                        if _determine_action_type(a) == _AnalyzeActionsType.RECOGNIZE_LINKED_ENTITIES
-                    ]
+                    a for a in generated_tasks
+                    if _determine_task_type(a) == _AnalyzeActionsType.RECOGNIZE_LINKED_ENTITIES
                 ],
                 sentiment_analysis_tasks=[
-                    t.to_generated() for t in
-                    [a for a in actions if _determine_action_type(a) == _AnalyzeActionsType.ANALYZE_SENTIMENT]
+                    a for a in generated_tasks if _determine_task_type(a) == _AnalyzeActionsType.ANALYZE_SENTIMENT
                 ]
             )
             analyze_body = self._client.models(api_version='v3.1').AnalyzeBatchInput(

sdk/textanalytics/azure-ai-textanalytics/azure/ai/textanalytics/aio/_text_analytics_client_async.py

@@ -17,7 +17,7 @@
 from azure.core.exceptions import HttpResponseError
 from azure.core.credentials import AzureKeyCredential
 from ._base_client_async import AsyncTextAnalyticsClientBase
-from .._request_handlers import _validate_input, _determine_action_type, _check_string_index_type_arg
+from .._request_handlers import _validate_input, _determine_task_type, _check_string_index_type_arg
 from .._response_handlers import (
     process_http_response_error,
     entities_result,
@@ -838,7 +838,7 @@ async def begin_analyze_actions(  # type: ignore
         :keyword bool show_stats: If set to true, response will contain document level statistics.
         :keyword int polling_interval: Waiting time between two polls for LRO operations
             if no Retry-After header is present. Defaults to 30 seconds.
-        :return: An instance of an LROPoller. Call `result()` on the poller
+        :return: An instance of an AsyncAnalyzeActionsLROPoller. Call `result()` on the poller
             object to return a pageable heterogeneous list of lists. This list of lists is first ordered
             by the documents you input, then ordered by the actions you input. For example,
             if you have documents input ["Hello", "world"], and actions
@@ -850,7 +850,7 @@ async def begin_analyze_actions(  # type: ignore
             Then, you will get the :class:`~azure.ai.textanalytics.RecognizeEntitiesResult` and
             :class:`~azure.ai.textanalytics.AnalyzeSentimentResult` of "world".
         :rtype:
-            ~azure.core.polling.AsyncLROPoller[~azure.core.async_paging.AsyncItemPaged[
+            ~azure.core.polling.AsyncAnalyzeActionsLROPoller[~azure.core.async_paging.AsyncItemPaged[
             list[
             RecognizeEntitiesResult or RecognizeLinkedEntitiesResult or RecognizePiiEntitiesResult or
             ExtractKeyPhrasesResult or AnalyzeSentimentResult
@@ -879,32 +879,26 @@ async def begin_analyze_actions(  # type: ignore
         continuation_token = kwargs.pop("continuation_token", None)
 
         doc_id_order = [doc.get("id") for doc in docs.documents]
-        task_order = [_determine_action_type(action) for action in actions]
+        generated_tasks = [action.to_generated(str(idx)) for idx, action in enumerate(actions)]
+        task_order = [(_determine_task_type(a), a.task_name) for a in generated_tasks]
 
         try:
             analyze_tasks = self._client.models(api_version='v3.1').JobManifestTasks(
                 entity_recognition_tasks=[
-                    t.to_generated() for t in
-                    [a for a in actions if _determine_action_type(a) == _AnalyzeActionsType.RECOGNIZE_ENTITIES]
+                    a for a in generated_tasks if _determine_task_type(a) == _AnalyzeActionsType.RECOGNIZE_ENTITIES
                 ],
                 entity_recognition_pii_tasks=[
-                    t.to_generated() for t in
-                    [a for a in actions if _determine_action_type(a) == _AnalyzeActionsType.RECOGNIZE_PII_ENTITIES]
+                    a for a in generated_tasks if _determine_task_type(a) == _AnalyzeActionsType.RECOGNIZE_PII_ENTITIES
                 ],
                 key_phrase_extraction_tasks=[
-                    t.to_generated() for t in
-                    [a for a in actions if _determine_action_type(a) == _AnalyzeActionsType.EXTRACT_KEY_PHRASES]
+                    a for a in generated_tasks if _determine_task_type(a) == _AnalyzeActionsType.EXTRACT_KEY_PHRASES
                 ],
                 entity_linking_tasks=[
-                    t.to_generated() for t in
-                    [
-                        a for a in actions if \
-                        _determine_action_type(a) == _AnalyzeActionsType.RECOGNIZE_LINKED_ENTITIES
-                    ]
+                    a for a in generated_tasks
+                    if _determine_task_type(a) == _AnalyzeActionsType.RECOGNIZE_LINKED_ENTITIES
                 ],
                 sentiment_analysis_tasks=[
-                    t.to_generated() for t in
-                    [a for a in actions if _determine_action_type(a) == _AnalyzeActionsType.ANALYZE_SENTIMENT]
+                    a for a in generated_tasks if _determine_task_type(a) == _AnalyzeActionsType.ANALYZE_SENTIMENT
                 ]
             )
             analyze_body = self._client.models(api_version='v3.1').AnalyzeBatchInput(

sdk/textanalytics/azure-ai-textanalytics/tests/test_analyze.py

@@ -679,3 +679,84 @@ def callback(resp):
             polling_interval=self._interval(),
             raw_response_hook=callback,
         ).result()
+
+    @GlobalTextAnalyticsAccountPreparer()
+    @TextAnalyticsClientPreparer()
+    def test_partial_success_for_actions(self, client):
+        docs = [{"id": "1", "language": "tr", "text": "I did not like the hotel we stayed at."},
+                {"id": "2", "language": "en", "text": "I did not like the hotel we stayed at."}]
+
+        response = client.begin_analyze_actions(
+                docs,
+                actions=[
+                    AnalyzeSentimentAction(),
+                    RecognizePiiEntitiesAction(),
+                ],
+                polling_interval=self._interval(),
+            ).result()
+
+        action_results = list(response)
+        assert len(action_results) == len(docs)
+        action_order = [
+            _AnalyzeActionsType.ANALYZE_SENTIMENT,
+            _AnalyzeActionsType.RECOGNIZE_PII_ENTITIES,
+        ]
+
+        assert len(action_results[0]) == len(action_order)
+        assert len(action_results[1]) == len(action_order)
+
+        # first doc
+        assert isinstance(action_results[0][0], AnalyzeSentimentResult)
+        assert action_results[0][0].id == "1"
+        assert action_results[0][1].is_error
+        assert action_results[0][1].id == "1"
+
+        # second doc
+        assert isinstance(action_results[1][0], AnalyzeSentimentResult)
+        assert action_results[1][0].id == "2"
+        assert isinstance(action_results[1][1], RecognizePiiEntitiesResult)
+        assert action_results[1][1].id == "2"
+
+    @pytest.mark.skip("Service bug - https://msazure.visualstudio.com/Cognitive%20Services/_workitems/edit/10145316")
+    @GlobalTextAnalyticsAccountPreparer()
+    @TextAnalyticsClientPreparer()
+    def test_multiple_of_same_action(self, client):
+        docs = [{"id": "1", "text": "My SSN is 859-98-0987."},
+                {"id": "2", "text": "Is 998.214.865-68 your Brazilian CPF number?"}]
+
+        response = client.begin_analyze_actions(
+                docs,
+                actions=[
+                    AnalyzeSentimentAction(),
+                    RecognizePiiEntitiesAction(),
+                    RecognizePiiEntitiesAction(domain_filter="phi"),
+                ],
+                polling_interval=self._interval(),
+            ).result()
+
+        action_results = list(response)
+        assert len(action_results) == len(docs)
+        action_order = [
+            _AnalyzeActionsType.ANALYZE_SENTIMENT,
+            _AnalyzeActionsType.RECOGNIZE_PII_ENTITIES,
+            _AnalyzeActionsType.RECOGNIZE_PII_ENTITIES,
+        ]
+
+        assert len(action_results[0]) == len(action_order)
+        assert len(action_results[1]) == len(action_order)
+
+        # first doc
+        assert isinstance(action_results[0][0], AnalyzeSentimentResult)
+        assert action_results[0][0].id == "1"
+        assert isinstance(action_results[0][1], RecognizePiiEntitiesResult)
+        assert action_results[0][1].id == "1"
+        assert isinstance(action_results[0][2], RecognizePiiEntitiesResult)
+        assert action_results[0][2].id == "1"
+
+        # second doc
+        assert isinstance(action_results[1][0], AnalyzeSentimentResult)
+        assert action_results[1][0].id == "2"
+        assert isinstance(action_results[1][1], RecognizePiiEntitiesResult)
+        assert action_results[1][1].id == "2"
+        assert isinstance(action_results[1][2], RecognizePiiEntitiesResult)
+        assert action_results[1][2].id == "2"